The Altair Community is migrating to a new platform to provide a better experience for you. The RapidMiner Community will merge with the Altair Community at the same time. In preparation for the migration, both communities are on read-only mode from July 15th - July 24th, 2024. Technical support via cases will continue to work as is. For any urgent requests from Students/Faculty members, please submit the form linked here.

stacking operator

ThiruThiru Member Posts: 100 Guru
dear all,  For my classification problem using supervised learning, Im trying to use stacking operator. 
The tutorial for this operator shows two layer, i.e. one base  learner layer + stacking  learner.   
My query is , can we have more than two layers...?
For example,  is it possible to have 3 layers,  i.e.  two layers of base learner , before I fed to the stacking learner?  
If so, how to configure that?  kindly let me know. thanks.


Best Answer

  • Options
    Telcontar120Telcontar120 Moderator, RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 1,635 Unicorn
    Solution Accepted
    Can't you just put one stacking operator inside another?  So your "base learner" of the outer stack would be another stacking operator of its own?
    Or in theory you could use any other ensemble operator as well.  Basically anything that gives you a model output which is all that the stacking operator requires as an input on the left side. 
    Brian T.
    Lindon Ventures 
    Data Science Consulting from Certified RapidMiner Experts


  • Options
    ThiruThiru Member Posts: 100 Guru
    thanks for your reply.  yes , it works. 

    1.  Is there any guideline how to combine the operators. 
    Using Gradient boost alone gives  accuracy 75.75%.  I want to take it beyond 95% using stacking.
     (ofcourse I'll look into precision/recall). 

     I tried many stacking combinations,  i couldnt make it beyond 73%.

    2. Any theoretical/white paper reference / rapidminer resource/ book on how to choose the  combination of  base learners , stack learner  for the maximum classification performance.  

    3.  will it be useful to use non-neural net operator as base learner & neural net (deep learning ) as stacking learner?
    i tried this combination, in my case, it reduces the performance to 55%.
    thank you

Sign In or Register to comment.